서브메뉴
검색
Advanced Machine Learning and Artificial Intelligence Methods for Spatial Omics
Advanced Machine Learning and Artificial Intelligence Methods for Spatial Omics
Detailed Information
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202103126
- ISBN
- 9798315729082
- DDC
- 574
- 저자명
- Chen, Jiawen.
- 서명/저자
- Advanced Machine Learning and Artificial Intelligence Methods for Spatial Omics
- 발행사항
- [Sl] : The University of North Carolina at Chapel Hill, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 160 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 86-11, Section: B.
- 주기사항
- Advisor: Li, Yun;Li, Didong.
- 학위논문주기
- Thesis (Ph.D.)--The University of North Carolina at Chapel Hill, 2025.
- 초록/해제
- 요약Spatial transcriptomics (ST) technology has revolutionized our understanding of biological systems by enabling the simultaneous measurement of gene expression and the preservation of spatial localization within tissues. Given this enriched data source, the central question in spatial omics analysis becomes: How can we efficiently utilize all this information and what important biological question can we answer? In this dissertation, we aim to develop ML and AI methods to address the pressing biological problems in spatial omics data analysis.In the first section, we present POLARIS, a versatile ST analysis method that can perform cell type deconvolution, identify anatomical or functional layer-wise differentially expressed (LDE) genes, and enable cell composition inference from histology images. POLARIS employs transfer learning to extract meaningful features from images of each spot and its surroundings. These features are then input into a deep neural network guided by Bayesian posterior probabilities to estimate cell compositions. This integration significantly enhances accuracy, and allows for predicting cell-type proportions in unmeasured regions and applying to new histology images without corresponding gene expression data.In the second section, we propose StarTrail, a novel ML method that leverage spatial gradients for spatial omics data. StarTrail investigates where and how does the omics feature change through spatial gradients, which map the rate of change of omics features across spatial positions. This approach allows us to identify zones of rapid change, characterized by high spatial gradients, which often correspond to critical biological junctures such as interfaces between lesional/diseased and healthy/normal tissues or regions with distinct functions. StarTrail, filling important gaps in current literature, enables deeper insights into tissue spatial architecture.In the third section, we introduce the Nearest Neighbor Derivative Process (NNDP), a scalable Gaussian Process framework that jointly models spatial processes and their derivatives, resulting a time complexity reduction from O(n3) to O(n). NNDP improves StarTrail in the second section by providing less hyper-parameter tuning, more theoretical support. NNDP robustly and accurately estimates spatial derivatives in various simulated spatial patterns and real data analysis.
- 일반주제명
- Biostatistics
- 일반주제명
- Cellular biology
- 일반주제명
- Bioinformatics
- 일반주제명
- Genetics
- 키워드
- Gene expression
- 키워드
- Cell type
- 키워드
- Histology images
- 키워드
- StarTrail
- 기타저자
- The University of North Carolina at Chapel Hill Biostatistics
- 기본자료저록
- Dissertations Abstracts International. 86-11B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017357068
■00520260202103126
■006m o d
■007cr#unu||||||||
■020 ▼a9798315729082
■035 ▼a(MiAaPQ)AAI31938728
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a574
■1001 ▼aChen, Jiawen.
■24510▼aAdvanced Machine Learning and Artificial Intelligence Methods for Spatial Omics
■260 ▼a[Sl]▼bThe University of North Carolina at Chapel Hill▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a160 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 86-11, Section: B.
■500 ▼aAdvisor: Li, Yun;Li, Didong.
■5021 ▼aThesis (Ph.D.)--The University of North Carolina at Chapel Hill, 2025.
■520 ▼aSpatial transcriptomics (ST) technology has revolutionized our understanding of biological systems by enabling the simultaneous measurement of gene expression and the preservation of spatial localization within tissues. Given this enriched data source, the central question in spatial omics analysis becomes: How can we efficiently utilize all this information and what important biological question can we answer? In this dissertation, we aim to develop ML and AI methods to address the pressing biological problems in spatial omics data analysis.In the first section, we present POLARIS, a versatile ST analysis method that can perform cell type deconvolution, identify anatomical or functional layer-wise differentially expressed (LDE) genes, and enable cell composition inference from histology images. POLARIS employs transfer learning to extract meaningful features from images of each spot and its surroundings. These features are then input into a deep neural network guided by Bayesian posterior probabilities to estimate cell compositions. This integration significantly enhances accuracy, and allows for predicting cell-type proportions in unmeasured regions and applying to new histology images without corresponding gene expression data.In the second section, we propose StarTrail, a novel ML method that leverage spatial gradients for spatial omics data. StarTrail investigates where and how does the omics feature change through spatial gradients, which map the rate of change of omics features across spatial positions. This approach allows us to identify zones of rapid change, characterized by high spatial gradients, which often correspond to critical biological junctures such as interfaces between lesional/diseased and healthy/normal tissues or regions with distinct functions. StarTrail, filling important gaps in current literature, enables deeper insights into tissue spatial architecture.In the third section, we introduce the Nearest Neighbor Derivative Process (NNDP), a scalable Gaussian Process framework that jointly models spatial processes and their derivatives, resulting a time complexity reduction from O(n3) to O(n). NNDP improves StarTrail in the second section by providing less hyper-parameter tuning, more theoretical support. NNDP robustly and accurately estimates spatial derivatives in various simulated spatial patterns and real data analysis.
■590 ▼aSchool code: 0153.
■650 4▼aBiostatistics
■650 4▼aCellular biology
■650 4▼aBioinformatics
■650 4▼aGenetics
■653 ▼aSpatial transcriptomics
■653 ▼aGene expression
■653 ▼aCell type
■653 ▼aHistology images
■653 ▼aStarTrail
■690 ▼a0308
■690 ▼a0379
■690 ▼a0369
■690 ▼a0715
■71020▼aThe University of North Carolina at Chapel Hill▼bBiostatistics.
■7730 ▼tDissertations Abstracts International▼g86-11B.
■790 ▼a0153
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17357068▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
Preview
Export
ChatGPT Discussion
AI Recommended Related Books
Подробнее информация.
- Бронирование
- не существует
- моя папка
- Первый запрос зрения
- Non-Book Loan Application
- Nighttime Book Loan Application
Available after logging in.


